ORIGINAL ARTICLE See related commentary on pg 1853

Activation of the IL-1b-Processing Is Involved in Contact Hypersensitivity Hideki Watanabe1,5, Olivier Gaide1,5, Virginie Pe´trilli2, Fabio Martinon2,6, Emmanuel Contassot1, Ste´phanie Roques1, Jean A. Kummer3,Ju¨rg Tschopp2 and Lars E. French4

The inflammasome is a cytosolic complex regulating the activation of -1, which cleaves the pro-inflammatory cytokines IL-1b and IL-18 into their active form. The inflammasome is composed of a NACHT-, LRR- and pyrin (NALP) family member that acts as a sensor for danger signals and the adaptor protein - associated speck-like protein containing a CARD domain (ASC), which allows the recruitment of caspase-1 in the complex. In the skin, exposure to contact sensitizers (CS) such as trinitro-chlorobenzene causes an immune response called contact hypersensitivity (CHS) or eczema. In this delayed-type hypersensitivity response, efficient priming of the adaptive immunity depends on the concomitant activation of the innate immune system, including IL-1b/IL-18 activation in the skin. To determine if the inflammasome contributes to CHS, we have analyzed its capacity to react to CS in vitro and in vivo. We show here that key components of the inflammasome are present in human keratinocytes and that CS like trinitro-chlorobenzene induce caspase-1/ ASC dependent IL-1b and IL-18 processing and secretion. We also show that ASC- and NALP3-deficient mice display an impaired response to CS. These findings suggest that CS act as danger signals that activate the inflammasome in the skin, and reveal a new role of NALP3 and ASC as regulators of innate immunity in CHS. Journal of Investigative Dermatology (2007) 127, 1956–1963; doi:10.1038/sj.jid.5700819; published online 12 April 2007

INTRODUCTION sensitivity (CHS) (Grabbe and Schwarz, 1998). There are few The skin protects us from our environment and is therefore T cells in normal skin and antigen presentation occurs mostly constantly subjected to microbial, chemical, and physical in draining lymph nodes. Uptake, transport, and presentation insults that ultimately result in the generation of non-self of antigens is achieved primarily by specialized skin cells antigens. Accordingly, the skin is endowed with the capacity called Langerhans cells (LCs) (Grabbe and Schwarz, 1998). to trigger immune responses (Murphy et al., 2000). Sensitiz- Efficient LC migration and antigen presentation as well as the ing chemicals that penetrate the skin surface induce a later recruitment of primed lymphocytes to the skin is delayed-type hypersensitivity reaction called contact hyper- dependent on the concomitant presence of pro-inflammatory stimuli (like IL-1b) induced by the contact sensitizers (CS) themselves (Grabbe et al., 1996; Griffiths et al., 2005). 1 Departments of Dermatology and Pathology/Immunology, Geneva However, how sensitizing chemicals induce activation of University Medical School, Geneva, Switzerland; 2Department of Biochemistry, University of Lausanne, Epalinges, Switzerland; 3Department of these cytokines is still unknown. Pathology, University Medical Center Utrecht, Utrecht, The Netherlands and IL-1b and IL-18 mediate acute inflammatory responses and 4Deptartment of Dermatology, Zurich University Hospital, Zu¨rich, provide a link between innate and adaptive immunity in CHS Switzerland (Shornick et al., 1996; Wang et al., 2002). Both cytokines are 5 These two authors contributed equally to this work. produced as inactive cytoplasmic precursors, which are 6 Current address: Harvard School of Public Health, Department of processed into their active forms by caspase-1 (Cerretti et al., Immunology and Infectious Diseases, Franc¸ois-Xavier Bagnoud Bldg; Room FXB205, 651 Huntington Ave, Boston MA 02115, USA. 1992; Thornberry et al., 1992; Ghayur et al., 1997; Dinarello, Correspondence: Professor Lars E. French, Department of Dermatology, 1998). Interestingly, a functional caspase-1 was found to be Zurich University Hospital, Gloriastrasse 31, CH-8091 Zu¨rich, Switzerland. necessary for LC migration and optimal CHS in mice E-mail: [email protected] or (Antonopoulos et al., 2001). The molecular mechanisms Dr Olivier Gaide, Department of Dermatology and Pathology/Immunology, controlling caspase-1 activity were totally unknown until Geneva University Medical School, Michel-Servet 1, CH-1211 Geneva, Switzerland. E-mail: [email protected] very recent studies demonstrated that it relied on a cytosolic Abbreviations: ASC, apoptosis-associated speck-like protein containing a multiprotein complex composed of a NACHT-, LRR-, and CARD domain; CHS, contact hypersensitivity; CS, contact sensitizers; (NALP) family member, the protein apoptosis- DAMPs, danger associated molecular patterns; LC, Langerhans cell; associated speck-like protein containing a CARD (ASC) and NALP, ACHT- (Neuronal apoptosis inhibitory protein, CIITA, HET-E, TP1), caspase-1 (Srinivasula et al., 2002; Agostini et al., 2004; LRR- and pyrin domain; PAMPs, pathogen associated molecular patterns; PK, primary keratinocytes; TNCB, trinitrochlorobenzene Martinon and Tschopp, 2005). As this complex regulates pro- Received 30 October 2006; revised 9 February 2007; accepted 15 February inflammatory cytokine activation, it was given the name 2007; published online 12 April 2007 inflammasome (Martinon and Tschopp, 2004).

1956 Journal of Investigative Dermatology (2007), Volume 127 & 2007 The Society for Investigative Dermatology H Watanabe et al. Contact Hypersensitivity and the Inflammasome

The role of the 14 human NALP identified is still a 4 hours 10 hours poorly understood. NALPs form a subfamily of the NACHT- LRR family that includes nucleotide oligomerisation domain

(NOD) proteins and is reminiscent of Toll-like receptors NT TLR-5 TLR-6 TNF TLR-5 TLR-6 TNF (Martinon and Tschopp, 2005). Interestingly, the latter are Pro-IL-1 key sensors of the innate immune system that detect pathogen-associated molecular patterns (Akira and Takeda, 140 2004). Recent studies using mice deficient in NALP3 or ASC b − show that NALPs may sense bacterial components and uric 120 zVAD acid crystals. Upon exposure to these ‘‘danger signals’’, they induce ASC-mediated caspase-1 recruitment and activation, 100 thereby triggering an inflammatory response (Kanneganti et al., 2006; Mariathasan et al., 2006; Martinon et al., 2006). 80 Hence, NALPs complete the NOD and toll-like receptors 60 (TLR) repertoire of danger sensors. IL-1 (pg/ml) Mutations in the encoding NALP3 are associated 40 with severe auto-inflammatory disorders characterized by recurrent fevers owing to excessive IL-1b secretion that are 20 alleviated by IL-1 receptor antagonists (Hawkins et al., 2003; Agostini et al., 2004; Hoffman et al., 2004). This underlines 0 the central role of the inflammasome in IL-1b regulation and the necessity to tightly control its activity. As patients affected TNCB NT NT DNFB DNCB SDS UVA UVB by these diseases have inflammatory skin lesions, we − hypothesized that an active inflammasome is present in the +TNF skin and that it may be responsible for CS-mediated IL-1b Figure 1. Danger signals induce caspase-dependent IL-1b secretion by activation. We show here that the inflammasome is indeed keratinocytes. (a) Stimulation of primary keratinocytes by NF-kB activators present and can be activated in keratinocytes. Moreover, such as TLR5/6 or TNFa (100 ng/ml) results in an increase in pro-IL-1b ASC- and NALP3-deficient mice display an impaired CHS to expression levels. (b) Exposure to TNFa itself does not result in a significant secretion of IL-1b, but TNFa-pretreated cells respond to TNCB (0.1, 1, and trinitro-chlorobenzene (TNCB) and dinitro-1-fluorobenzene 5 mg/ml), SDS (15 mg/ml), or UVB (80 mJ/cm2) for 24 hours by secreting IL-1b (DNFB), thus identifying the NALP3 inflammasome as a key in the supernatant (ELISA detection, nX6), as opposed to stimulation by regulator of innate immunity in contact hypersensitivity. UVA (80 mJ/cm2) exposure. In each case, IL-1b secretion is blocked by the Hence, pro-inflammatory signals necessary for the priming of pan-caspase inhibitor zVAD (20 mM). T cells are provided, at least in part, by danger sensors of the NALP family. to have a similar effect on cultured keratinocytes. This RESULTS increase in pro-IL-1b was not paralleled by a significant rise TNCB, SDS, and UV induce caspase-dependent IL-1b secretion in IL-1b secretion (Figure 1b). However, significant secretion by keratinocytes of IL-1b could be induced by exposure to the CS TNCB, Evidence suggests that several cytokines implicated in CHS DNFB, and dinitro-1-chlorobenzene. Likewise, SDS and UVB are produced by resident cells within the skin (Grabbe et al., significantly increased IL-1b secretion, whereas UVA induced 1996; Griffiths et al., 2005). In particular, LC and keratino- only a minor increase in IL-1b levels. IL-1b secretion was cytes can secrete IL-1b and IL-18 upon exposure to CS (Enk dependent on caspase activity, as it was blocked by the pan- et al., 1993; Naik et al., 1999) and these cytokines are crucial caspase inhibitor benzyloxycarbonyl-valine-alanine-aspartate for LC migration to the lymph nodes (Cumberbatch et al., (zVAD) (Figure 1b). Apoptosis was not observed at these 2002). We decided to determine if keratinocytes, which form doses of TNCB and UV, suggesting that caspase-1 may be the major cell population in the skin, could react to danger involved. In accordance with this, targeted overexpression of signals by activating and secreting IL-1b. In , caspase-1 in mouse keratinocytes was shown to result in CHS IL-1b secretion follows two sequential events: (i) upregulation symptoms dependent on IL-1b and IL-18 (Yamanaka et al., of pro-IL-1b through NF-kB-mediated transcriptional activa- 2000). Keratinocytes can thus sense danger signals and tion (Martinon and Tschopp, 2005) and (ii) inflammasome/ subsequently activate IL-1b independently of the immune caspase-1-mediated processing of pro-IL-1b (Dinarello, 1998; system. Although TNF prestimulation allowed an easier Martinon et al., 2002). We found that exposure of primary detection of IL-1b, it was not essential for its activation and keratinocytes to purified flagellin (activating TLR5), MALP-2 secretion (Figure 3a). (activating TLR6), or TNFa, which all activate NF-kB indeed resulted in an increase of the expression level of pro-IL-1b Both keratinocytes and monocyte-derived cells contain (Figure 1a and data not shown). The promoter of pro-IL-1b molecular components required for inflammasome assembly is known to contain an NF-kB binding site (Hiscott et al., In macrophages, pro-IL-1b processing is dependent on the 1993), and other potent NF-kB activators are therefore likely assembly of the inflammasome, composed of a NALP protein,

www.jidonline.org 1957 H Watanabe et al. Contact Hypersensitivity and the Inflammasome

the adaptor protein ASC, and caspase-1, which cleaves it secretion as previously reported (Dinarello, 1998; Martinon directly (Dinarello, 1998; Martinon et al., 2002). We et al., 2002), but IL-18 activation was unchanged. This may reasoned that resident skin cells may use the same machinery be explained by the observation that both caspase-1 and and assessed the presence of molecular components of the pro-IL-1b expression levels are increased by TNFa but not inflammasome in keratinocytes and Langerhans-like cells. IL-18 (Figure 4a, cell extracts). Messenger RNA coding for all NALPs 1–6, 8, 10–11, and TNCB-induced caspase-1 cleavage resulted in a decrease 13–14 were detectable in primary keratinocytes and human of the intracellular pool of full-length caspase-1 (Figure 3a, adult calcium-reduced temperature elevated keratinocyte cell extracts). Shorter fragments of capase-1 were not (HaCaT) cells (Figure 2a). NALP7 was only detectable in detected in the cell extracts, which is consistent with previous primary keratinocytes, whereas NALP12 could only be dete- reports showing that cleaved caspase-1 is secreted together cted in HaCaT cells. Messenger RNA for all NALPs except with IL-1b (Martinon et al., 2002). The kinetic of caspase-1 NALP 5, 7, and 9 were observed in Tohoku University, (p35 and p11) and IL-1b activation, appearing 6 hours after Department of Pediatrics-1 (THP-1) cells, a monocyte- stimulation and increasing thereafter (Figure 3b and data not derived cell line. Monocytes have the potential to differenti- shown), was similar in monocyte-derived cells stimulated ate into dendritic cells and respond to sensitizers, thus with uric acid crystals (Martinon et al., 2006). Concomitant resembling LCs (Ashikaga et al., 2002). Messenger RNA for exposure to the caspase-1-specific inhibitor benzyloxycarbo- ASC-1, caspase-1, caspase-5, pro-IL-1b, and IL-18 were nyl-tyrosine-valine-alanine-aspartate (zYVAD) blocked both detectable in the three cell types (Figure 2b). caspase-1 and IL-1b activation and secretion (Figure 3b). Expression at the protein level was confirmed by SDS- Hence, TNCB-induced caspase-1 activation in keratinocytes PAGE/Western blot analysis of whole-cell proteins extracts. is necessary for IL-1b activation. NALP1 and NALP3, ASC, caspase-1 and caspase-5 could be We confirmed the central role of the inflammasome by readily detected in primary keratinocytes, HaCaT cells, and isolating mouse keratinocytes from ASC-deficient mice and THP-1 cells (Figure 2c). Levels of NALP1, NALP3, and exposing them to TNCB. ASC-deficient keratinocytes failed to caspase-1 were lower in keratinocytes than in THP-1 cells, secrete IL-1b upon in vitro exposure to TNCB, in contrast to whereas levels of caspase-5 and ASC were similar. This keratinocytes isolated from wild-type mice (Figure 3c). suggests that resident skin cells contain the molecular Hence, TNCB-induced IL-1b activation and secretion by components necessary for inflammasome assembly. keratinocytes is likely to be controlled by the inflammasome.

TNCB induces ASC-dependent caspase-1 and IL-1b activation Contact hypersensitivity to DNFB and TNCB is reduced in keratinocytes in ASC and NALP3-deficient mice We next looked at the role of caspase-1 in TNCB-induced IL- In order to confirm the role of the inflammasome in vivo,we 1b activation by exposing primary keratinocytes to TNCB and used a model of CHS to DNFB, where 6- to 7-week-old ASC- analyzing the cellular and secreted caspase-1/IL-1b patterns. deficient mice and their wild-type littermates were sensitized Activated (cleaved) fragments of caspase-1 (p35 and p11), IL- either on the left ear or on their shaved belly and then 18 (p18), and IL-1b (p17) could be found in supernatants only challenged on day 5 at a distant site (i.e. the right ear). Ear after TNCB stimulation (Figure 3a, supernatants). Pretreat- thickness measured 24 hours after the challenge showed a ment with TNFa increased capase-1 cleavage and IL-1b significant diminution of the ear swelling in ASC-deficient

abc - PK - HaCaT - THP-1 - PK - HaCaT - THP-1 - PK - HaCaT - THP-1

NALP1 NALP8 ASC - THP-1 - HaCaT - PK - Mock - Transfected

NALP2 NALP9 Casp-1 NALP-1

NALP3 NALP10 NALP-3 Casp-5 NALP4 NALP11 Casp-1 IL-1 NALP5 NALP12 Casp-5 IL-18 NALP6 NALP13 ASC

NALP7 NALP14

Figure 2. Keratinocytes and monocyte-derived cells contain the molecular components necessary for inflammasome activity. (a) Reverse transcription-PCR analysis shows detectable levels of mRNA coding for most members of the NALP family in primary human keratinocytes (PK), keratinocyte-derived HaCaT cells, and the monocyte-derived THP-1 cell line. (b) Messenger RNA of ASC, caspase-1, caspase 5, IL1-b, and IL-18 are present in primary keratinocytes (PK), HaCaT, and THP-1 cells. (c) Western blot analysis reveals the presence of NALP1, NALP3, caspase-1, caspase-5, and ASC at the protein level in PK, HaCaT, and THP-1 cells. The specificity of each antibody was confirmed by Western blot analysis of extracts from 293T cells transfected or not with the relevant complementary DNAs. The data presented are representative of at least three independent experiments.

1958 Journal of Investigative Dermatology (2007), Volume 127 H Watanabe et al. Contact Hypersensitivity and the Inflammasome

a NT TNCB b NT TNCB +zYVAD 24 6 12 24 6 12 24 80 6 hours

- PBS - TNF - PBS - TNF 12 hours 240 60 24 hours 160 40

80 IL-1 (pg/ml) 20 0 0 SN Casp-1 (fl) SN Casp-1 (fl) Casp-1 (p35) Casp-1 (p35)

Casp-1 (p11) Casp-1 (p11) c IL-18 (p18) NT TNCB

IL-1 (p17)

CE IL-1 (p17) - ASC +/+ - ASC − / - ASC +/+ - ASC − / pro-IL-1 SN IL-1 (p17) Casp-1 CE ASC

Figure 3. TNCB induces ASC-dependent caspase-1 and IL-1b activation in keratinocytes. (a) Primary keratinocytes exposed to TNCB (5 mg/ml) secrete active caspase-1 (p35 and p11), IL-18 (p18), and IL-1b (p17), as revealed by Western blot and ELISA analysis of supernatants (SN). This is paralleled by a decrease in full-length caspase-1 found in cell extracts (CE). TNFa increases the expression levels of caspase-1 and pro-IL-1b but not IL-18 (CE). TNFa pretreatment increases activation and secretion of both caspase-1 and IL-1b SN(b) Secretion of active caspase-1 (p35 and p11) in the supernatant (SN) and IL-1b (ELISA) starts 6 hours after TNCB exposure and is blocked by the caspase-1 inhibitor zYVAD (50 mM). (c) TNCB (1 mg/ml)-induced secretion of active IL-1b is reduced in primary keratinocytes cultures derived from neonatal ASC-deficient (/) mice as opposed to wild-type (WT) mice. mice (Figure 4a). In non-sensitized mice, DNFB failed to In ASC- and NALP3-deficient animals, TNCB-induced induce a detectable ear swelling, as previously described swelling was significantly lower in the early phase of (Riemann et al., 2005). The limited ear swelling observed elicitation (Figure 4c). The extent of this reduction was after sensitization on the belly may be due to the ingestion (by similar in ASC, NALP3, and IL-1R-deficient mice, although licking) of small amounts of DNFB, which acts as a tolerizer strain differences (BALB/c, DBA/C57BL/6, and C57BL/6, when applied orally (Faria and Weiner, 2005). respectively) resulted in different peak levels in ear swellings, Contact hypersensitivity (CHS) involves priming of naı¨ve as previously reported (Bigby et al., 1987). The reduction of T cells (sensitization) and activation of primed T cells upon the ear swelling was confirmed by histological analysis of re-exposure to the antigen (elicitation). Elicitation can be ASC and NALP3 deficient mice ears 5 hours after TNCB subdivided into an early and a late phase (Grabbe and application, which revealed a marked reduction in the Schwarz, 1998). The early phase, characterized by increased swelling, subcutaneous edema, and vasodilatation that vessel permeability and local edema, peaks 8 hours after normally characterize the early phase (Figure 5). The late antigen re-exposure and is believed to be mediated by local phase of elicitation, characterized by the recruitment of release of cytokines (including IL-1b) and histamine (Kitagaki inflammatory cells in the skin, was not significantly different et al., 1997; Grabbe and Schwarz, 1998). The late phase in ASC- and NALP3-deficient mice, suggesting that the develops 12–36hours after antigen re-exposure and is due to inflammasome only plays a minor role at this stage (Figure cellular infiltration. 4c and 5). We examined both the early and late phases of TNCB- We found that IL-1R deficiency affected the early phase as induced CHS in IL-1R-, ASC-, and NALP3-deficient mice. well as the late phase of elicitation, although the difference Mice 6–7 week old were sensitized twice to TNCB by observed was not always statistically significant (Figure 4b external application of TNCB on the ear at day 0 and 7, and and c). This either suggests that IL-1a in involved in the late ear thickness was measured 1, 2, 3.5, 5, 8, 14, 24, 32, and phase of elicitation (both IL-1a and b signal thought the IL-1R) 48 hours after elicitation with TNCB at day 14. Consistent or that IL-1b can be cleaved by other proteases, such as mast with previous reports, we observed a peak ear swelling cells chymases (Mizutani et al., 1991). However, activation 5–8hours after TNCB application (early phase), followed by a of caspase-1 was demonstrated to be essential for optimal second peak at 24 hours (late phase) (Figure 4b). IL-1R defi- sensitization in mice, suggesting that other proteases only ciency did not alter the kinetics, but resulted in a significant play a minor role in IL-1b activation in CHS (Antonopoulos decrease in the intensity of the ear swelling (Figure 4b and c). et al., 2001).

www.jidonline.org 1959 H Watanabe et al. Contact Hypersensitivity and the Inflammasome

a Opposite ear Shaved belly c sensitization sensitization ASC NALP3 IL-1R P=0.01 P=0.02 P=0.002 P=0.043 P=0.02 P=0.045 P=0.014 16 8 25 25 25 20 20 20 12 6 15 15 15 8 NS 4 10 10 10 2 4 5 5 5 ES at 24 hours (10 m)

0 early phase (10 m) ES: 0 0 +/+ −/− 0 +/+ −/− 0 +/+ −/− ASC: +/++/++/+ −/− +/+ +/+ −/− NS NS P=0.056

20 20 20 AOO–AOO

AOO–DNFB AOO–DNFB 15 15 15 DNFB–DNFB DNFB–DNFB DNFB–DNFB DNFB–DNFB

+/+ 10 10 10 b 20 IL-1R ; vehicle IL-1R +/+ ; TNCB 5 5 5 IL-1R −/− ; TNCB 0 0 0 ES: late phase (10 m) ES: +/+ −/− +/+ −/− +/+ −/− 10 * *

* * *

Ear swelling (10 m) Ear swelling * 0 0 2468 1020304050 Time after elicitation (hours)

Figure 4. Contact hypersensitivity to DNFB and TNCB is reduced in ASC-, NALP3-, and IL-1R-deficient mice. (a) ASC deficiency blocks the ear swelling measured 24 hours after challenge (day 5) with DNFB on the right ear of mice previously sensitized at distant sites (left ear or shaved belly) at days 0 and1(n ¼ 5 per group). Non-sensitized wild-type mice show barely discernable ear swelling after DNFB challenge and the vehicle (acetone:olive oil ¼ AOO) alone has no detectable effect either. (b–c) Ear swelling measured after elicitation (day 14) of TNCB-sensitized mice (days 0 and 7) is reduced during both the early and late phases of CHS in IL1R-deficient (/, filled triangles) mice as compared with wild-type ( þ / þ , filled circles) littermates. Asterisks denote statistical significance (*Po0.05; n ¼ 5 mice per group). (c) Ear swelling measured after elicitation of TNCB-sensitized mice is significantly reduced during the early phase of elicitation in ASC- and NALP3-deficient mice as compared with wild-type littermates, but not during the late phase. The extent of reduction of the ear swelling in the early phase of elicitation is similar in ASC-, NALP3-, and IL-1R-deficient mice, although strain differences result in different peak levels in ear swellings.

WT ASC −/− DISCUSSION We provide evidence that primary human keratinocytes NT contain the necessary elements to form an inflammasome and that TNCB induces caspase-1-mediated cleavage and activation of IL-1b in an ASC-dependent manner. The kinetics of the activation of caspase-1 and IL-1b cleavage and secretion was similar to the other inflammasome activators. TNCB can thus be considered as a new ‘‘danger signal’’ 5 hours sensed by the inflammasome, together with uric acid crystals and ATP (Mariathasan et al., 2006; Martinon et al., 2006). Interestingly, we found that other stimuli such as chemical irritants (SDS) and physical agents (UVB) could also trigger inflammasome activation, giving further support to the hypothesis that mammals evolved a common set of sensors for various danger signals, ranging form infections to physical 24 hours insults (Matzinger, 2002; Shi et al., 2003). It is unclear whether the capacity to sense CS is a means of defense/alarm or whether sensitizers ‘‘highjack’’ systems that have evolved to sense other target molecules such as bacterial products. In any case, inflammasome inhibitors may have the ability to Figure 5. Contact hypersensitivity to TNCB is reduced in ASC-deficient prevent inflammatory reactions owing to CS. mice. Histology of ASC-deficient and wild-type (wt) mice ears 5 hours after The importance of NALP3 and ASC in the early phase, and TNCB challenge (bar ¼ 200 mm). Edema and vasodilatation characteristic of the early phase of elicitation of CHS are less prominent in ASC-deficient not in the late phase, of elicitation suggests that the (/) mice as compared with wild-type littermates ( þ / þ ). No apparent inflammasome has little impact on the recruitment of primed histological difference is apparent 24 hours after elicitation with TNCB. T cells, but rather initiates pro-inflammatory signals that

1960 Journal of Investigative Dermatology (2007), Volume 127 H Watanabe et al. Contact Hypersensitivity and the Inflammasome

promote T-cell priming. This is consistent with a recent report factor, and 10% fetal calf serum. THP-1 cells were cultured in RPMI- showing that NALP3 is essential for the sensitization phase of 1640 medium (Invitrogen, NV Leeks, The Netherlands) containing contact hypersensitivity to trinitrophenylchloride (Sutterwala 2 105 Mofb-mercapto-ethanol and 10% fetal calf serum. Cell et al., 2006). cultures were maintained at 371C in humidified incubators with 5%

Low doses of CS are sufficient to stimulate pre-primed CO2. All media contained 100 U/ml of penicillin and 10 mg/ml of T cells but insufficient to elicit a CHS response (Grabbe et al., streptomycin. Purified flagellin (activating TLR5) and MALP-2 1996), suggesting that in addition to hapten-specific recogni- (activating TLR6) from the TLR activating kit (Alexis, Lausen, tion, a pro-inflammatory stimulus that is provided locally by Switzerland) were used at 2 mg/ml. TNFa (human recombinant the contact sensitizer itself is necessary (Grabbe et al., 1996). TNFa, Alexis, Lausen, Switzerland) was used at 1 mg/ml. zVAD and We propose that this pro-inflammatory stimulus is due to zYVAD (Alexis, Lausen, Switzerland) were used at a final concen- NALP3 inflammasome-dependent IL-1b production in the tration of 20 mM each. Mouse primary keratinocyte were obtained skin. However, other cytokines/signals are likely to modulate from 2-day-old mice as described (Pirrone et al., 2005). Briefly, the early phases of CHS, as IL-1R-deficiency did not totally whole skins were floated overnight in 0.25% trypsin at 41C. After abrogate ear swelling after elicitation with TNCB. This may mechanical separation from the dermis, the epidermis was be due at least in part to the presence of IL-18, which is also dissociated by 50 shakes in a 15-ml tube in 5 ml of medium. activated by caspase-1 (Wang et al., 2002; Kawase et al., Keratinocytes were seeded in 50% low calcium fibroblast-condi- 2003). tioned medium (EMEM, Cambrex, Verviers, Belgium) and preincu- Our data suggest that keratinocytes stimulated by CS may bated for 24 hours on dermal cells and 50% Epilife (Biocoba, Basel, directly provide the pro-inflammatory stimuli required for Switzerland) on collagen IV-coated dishes (non-keratinocyte will not CHS. However, several other cells such as LC, dermal adhere/grow in these conditions) for 7 days before stimulation with dendritic cells, and mast cells were also reported to play a TNCB or DNFB or 4,6-dinitro-1-chlorobenzene at 1–10 mg/ml. role in CHS (Kitagaki et al., 1997; Griffiths et al., 2005; Kaplan et al., 2005), and inflammasome activity in these cells Reverse transcription-PCR also be implicated. Interestingly, NOD2, which plays an Total RNA extracted from fresh THP-1, HaCaT, and primary essential role in inflammatory bowel disease, is expressed at keratinocytes using TRIzol reagent (Invitrogen, LV Leeks, The low levels in normal enterocytes, but is induced by TNFa in Netherlands) was reverse-transcribed and amplified using OneStep inflammatory conditions (Rosenstiel et al., 2003). We found RT-PCR kit (Qiagen, Basel, Switzerland) according to the manufac- likewise that TNFa significantly increases some of the turer’s instructions. Sequences of the oligonucleotide used for RT- inflammasome components in keratinocyte. This increase PCR are provided online in Table S1. was neither sufficient nor essential for IL-1b activation and secretion, although it potentiated the effect of CS. Interest- Western blot analysis and ELISA ingly, TNFa has been demonstrated to play a crucial role in Cell lysates and supernatants were subjected to SDS-PAGE/Western the cross-talk between LC and keratinocytes that drives LC blotting as described (Agostini et al., 2004). Primary antibodies: migration (Cumberbatch et al., 2003; Griffiths et al., 2005). human cleaved IL-1b (D116 epitope, Cell Signaling, Danvers), Taken together, these data suggest an active role of epithelial human proIL-1b (R&D Systems, Abingdon, UK), human ASC/Pycard, cells, enterocytes, and keratinocytes in the surveillance of NALP1 and 3 (Apotech, Epalinges, Swizerland), human caspase-1 danger signals and the activation of innate immunity signals. (Santa Cruz Biotechnology, LabForce AG, Nunningen, Switzerland), Blocking of NF-kB signaling using topical decoy oligonu- or human caspase-5 (MBL, LabForce AG, Nunningen, Switzerland). cleotide was recently reported to diminish symptoms of Secondary antibodies: horseradish peroxidase-conjugated anti-rabbit atopic dermatitis in a mouse model of this disease (Dajee or anti-mouse IgG (Jackson immunoresearch, Milan Analytica, La et al., 2006). Although this effect may not be related to a Roche, Switzerland). Detection: ECL (Amersham, Zurich, Switzer- decrease in pro-IL-1b levels, it suggests that targeting the land) as described in the manufacturer’s instructions. ELISA: inflammasome with small-molecule inhibitors (which remain undiluted cell culture supernatants were subjected to human IL-1b to be discovered) may represent a new modality for the detection by ELISA (BD biosciences, Basel, Switzerland) according treatment of CHS. The characterization of the role of the to the manufacturer’s instructions. inflammasome in inflammatory skin disorders must be therefore be pursued as it may eventually lead to develop- Analysis of CHS in IL-1R, ASC, and NALP3 mice ment of new treatment modalities for inflammatory skin Mice experiments were approved by the Geneva Cantonal diseases. Veterinary Office before being conducted, according to institutional and Swiss Federal Veterinary Office guidelines. IL-1R- and ASC- MATERIALS AND METHODS deficient mice were obtained from M Kopf (Molecular Biomedicine Cell culture ETH, Zurich, Switzerland) and VM Dixit (Genentech, San Francisco, Primary keratinocytes (HEK, Biocoba, Basel, Switzerland) were CA). NALP3-deficient mice were generated as described (Martinon cultured in Epilife supplemented with the provided human et al., 2006). Mice (7–8 week old) (n ¼ 5 per group) were sensitized þþ keratinocyte growth supplement and 60 mM of Ca and HaCaT by external applications of 20 ml of 0.5% DNFB in acetone:olive oil in DMEM:F-12 (1:1) (Invitrogen, NV Leeks, The Netherlands) (3:1, v/v) on the left ear or on their shaved belly at days 0 and 1. At supplemented with 5 mg/ml human insulin, 10 ng/ml cholera toxin, day 5, sensitized mice were elicited on the ventral and dorsal side of 0.4 mg/ml hydrocortisone, 10 ng/ml of human epidermal growth the right ear by topical application of 10 ml of 0.3% DNFB in

www.jidonline.org 1961 H Watanabe et al. Contact Hypersensitivity and the Inflammasome

acetone:olive oil. Ear thickness was measured 24 hours after Ghayur T, Banerjee S, Hugunin M, Butler D, Herzog L, Carter A et al. (1997) elicitation with a digital gauge (Mitsutoyo, Kanagawa, Japan) on Caspase-1 processes IFN-gamma-inducing factor and regulates LPS- induced IFN-gamma production. Nature 386:619–23 both ears. Alternatively, 6- to 7-week-old mice (X3 per group) were sensitized by external applications of 20 ml of 1% TNCB in acetone Grabbe S, Schwarz T (1998) Immunoregulatory mechanisms involved in elicitation of allergic contact hypersensitivity. Immunol Today on the right ear at days 0 and 7. On day 14, ear thickness was 19:37–44 measured before and after TNCB application (20 ml of 1% TNCB in Grabbe S, Steinert M, Mahnke K, Schwartz A, Luger TA, Schwarz T (1996) acetone) to the same ear. Dissection of antigenic and irritative effects of epicutaneously applied haptens in mice. Evidence that not the antigenic component but Statistical analysis nonspecific pro-inflammatory effects of haptens determine the concen- tration-dependent elicitation of allergic contact dermatitis. J Clin Invest Groups were compared using one-tailed Student’s or Aspin–Welch’s 98:1158–64 t-tests. Griffiths CE, Dearman RJ, Cumberbatch M, Kimber I (2005) Cytokines and langerhans cell mobilisation in mouse and man. Cytokine 32: CONFLICT OF INTEREST 67–70 The authors state no conflict of interest. Hawkins PN, Lachmann HJ, McDermott MF (2003) Interleukin-1-receptor antagonist in the Muckle–Wells syndrome. N Engl J Med 348: 2583–4 ACKNOWLEDGMENTS This work was supported by grants from the Swiss National Science Hiscott J, Marois J, Garoufalis J, D’Addario M, Roulston A, Kwan I et al. (1993) Foundation, the Louis-Jeantet Foundation, and the Commission of Technology Characterization of a functional NF-kappa B site in the human and Innovation (CTI). HW was supported by Mahuro Co Ltd, Japan and VP interleukin 1 beta promoter: evidence for a positive autoregulatory loop. was supported by a fellowship of the FRM (Fondation pour la Recherche Mol Cell Biol 13:6231–40 Medicale). Many thanks to Konrad Hauser for a critical reading of this paper. Hoffman HM, Rosengren S, Boyle DL, Cho JY, Nayar J, Mueller JL et al. (2004) Prevention of cold-associated acute inflammation in familial cold autoinflammatory syndrome by interleukin-1 receptor antagonist. Lancet SUPPLEMENTARY MATERIAL 364:1779–85 Table S1. Oligos used for reverse transcription–PCR detection of inflam- Kanneganti TD, Ozoren N, Body-Malapel M, Amer A, Park JH, Franchi L et al. masome components. (2006) Bacterial RNA and small antiviral compounds activate caspase-1 through cryopyrin/Nalp3. Nature 440:233–6 REFERENCES Kaplan DH, Jenison MC, Saeland S, Shlomchik WD, Shlomchik MJ (2005) Agostini L, Martinon F, Burns K, McDermott MF, Hawkins PN, Tschopp J Epidermal langerhans cell-deficient mice develop enhanced contact (2004) NALP3 forms an IL-1beta-processing inflammasome with hypersensitivity. Immunity 23:611–20 increased activity in Muckle–Wells autoinflammatory disorder. Immunity Kawase Y, Hoshino T, Yokota K, Kuzuhara A, Kirii Y, Nishiwaki E et al. (2003) 20:319–25 Exacerbated and prolonged allergic and non-allergic inflammatory Akira S, Takeda K (2004) Toll-like receptor signalling. Nat Rev Immunol cutaneous reaction in mice with targeted interleukin-18 expression in 4:499–511 the skin. J Invest Dermatol 121:502–9 Antonopoulos C, Cumberbatch M, Dearman RJ, Daniel RJ, Kimber I, Groves Kitagaki H, Ono N, Hayakawa K, Kitazawa T, Watanabe K, Shiohara T (1997) RW (2001) Functional caspase-1 is required for Langerhans cell Repeated elicitation of contact hypersensitivity induces a shift in migration and optimal contact sensitization in mice. J Immunol cutaneous cytokine milieu from a T helper cell type 1 to a T helper 166:3672–7 cell type 2 profile. J Immunol 159:2484–91 Ashikaga T, Hoya M, Itagaki H, Katsumura Y, Aiba S (2002) Evaluation of Mariathasan S, Weiss DS, Newton K, McBride J, O’Rourke K, Roose-Girma M CD86 expression and MHC class II molecule internalization in THP-1 et al. (2006) Cryopyrin activates the inflammasome in response to toxins human monocyte cells as predictive endpoints for contact sensitizers. and ATP. Nature 440:228–32 Toxicol In Vitro 16:711–6 Martinon F, Burns K, Tschopp J (2002) The inflammasome: a molecular Bigby M, Kwan T, Sy MS (1987) Ratio of Langerhans cells to Thy-1+ dendritic platform triggering activation of inflammatory and processing of epidermal cells in murine epidermis influences the intensity of contact proIL-beta. Mol Cell 10:417–26 hypersensitivity. J Invest Dermatol 89:495–9 Martinon F, Petrilli V, Mayor A, Tardivel A, Tschopp J (2006) Gout-associated Cerretti DP, Kozlosky CJ, Mosley B, Nelson N, Van Ness K, Greenstreet TA uric acid crystals activate the NALP3 inflammasome. Nature 440: et al. (1992) Molecular cloning of the interleukin-1 beta converting 237–41 enzyme. Science 256:97–100 Martinon F, Tschopp J (2004) Inflammatory caspases: linking an intracellular Cumberbatch M, Bhushan M, Dearman RJ, Kimber I, Griffiths CE (2003) IL- innate immune system to autoinflammatory diseases. Cell 117:561–74 1beta-induced Langerhans’ cell migration and TNF-alpha production in Martinon F, Tschopp J (2005) NLRs join TLRs as innate sensors of pathogens. human skin: regulation by lactoferrin. Clin Exp Immunol 132:352–9 Trends Immunol 26:447–54 Cumberbatch M, Dearman RJ, Groves RW, Antonopoulos C, Kimber I (2002) Matzinger P (2002) The danger model: a renewed sense of self. Science Differential regulation of epidermal langerhans cell migration by 296:301–5 interleukins (IL)-1alpha and IL-1beta during irritant- and allergen- induced cutaneous immune responses. Toxicol Appl Pharmacol Mizutani H, Schechter N, Lazarus G, Black RA, Kupper TS (1991) Rapid 182:126–35 and specific conversion of precursor interleukin 1 beta (IL-1 beta) to an active IL-1 species by human mast cell chymase. J Exp Med Dajee M, Muchamuel T, Schryver B, Oo A, Alleman-Sposeto J, De Vry CG 174:821–5 et al. (2006) Blockade of experimental atopic dermatitis via topical NF- kappa B decoy oligonucleotide. J Invest Dermatol 126:1792–803 Murphy JE, Robert C, Kupper TS (2000) Interleukin-1 and cutaneous inflammation: a crucial link between innate and acquired immunity. Dinarello CA (1998) Interleukin-1 beta, interleukin-18, and the interleukin-1 J Invest Dermatol 114:602–8 beta converting enzyme. Ann NY Acad Sci 856:1–11 Naik SM, Cannon G, Burbach GJ, Singh SR, Swerlick RA, Wilcox JN et al. Enk AH, Angeloni VL, Udey MC, Katz SI (1993) An essential role for (1999) Human keratinocytes constitutively express interleukin-18 and Langerhans cell-derived IL-1 beta in the initiation of primary immune secrete biologically active interleukin-18 after treatment with pro- responses in skin. J Immunol 150:3698–704 inflammatory mediators and dinitrochlorobenzene. J Invest Dermatol Faria AM, Weiner HL (2005) Oral tolerance. Immunol Rev 206:232–59 113:766–72

1962 Journal of Investigative Dermatology (2007), Volume 127 H Watanabe et al. Contact Hypersensitivity and the Inflammasome

Pirrone A, Hager B, Fleckman P (2005) Primary mouse keratinocyte culture. Srinivasula SM, Poyet JL, Razmara M, Datta P, Zhang Z, Alnemri ES (2002) In: Epidermal Cells: Methods and Protocols. (K. Turksen ed). Humana The PYRIN-CARD protein ASC is an activating adaptor for caspase-1. Press Inc: Totowa, 3–14 J Biol Chem 277:21119–22 Riemann H, Loser K, Beissert S, Fujita M, Schwarz A, Schwarz T et al. (2005) Sutterwala FS, Ogura Y, Szczepanik M, Lara-Tejero M, Lichtenberger GS, IL-12 breaks dinitrothiocyanobenzene (DNTB)-mediated tolerance and Grant EP et al. (2006) Critical role for NALP3/CIAS1/Cryopyrin in innate converts the tolerogen DNTB into an immunogen. J Immunol 175: and adaptive immunity through its regulation of caspase-1. Immunity 5866–74 24:317–27 Rosenstiel P, Fantini M, Brautigam K, Kuhbacher T, Waetzig GH, Seegert D Thornberry NA, Bull HG, Calaycay JR, Chapman KT, Howard AD, Kostura MJ et al. (2003) TNF-alpha and IFN-gamma regulate the expression of the et al. (1992) A novel heterodimeric cysteine protease is required for NOD2 (CARD15) gene in human intestinal epithelial cells. Gastro- interleukin-1 beta processing in monocytes. Nature 356:768–74 enterology 124:1001–9 Wang B, Feliciani C, Howell BG, Freed I, Cai Q, Watanabe H et al. (2002) Shi Y, Evans JE, Rock KL (2003) Molecular identification of a danger signal Contribution of Langerhans cell-derived IL-18 to contact hypersensiti- that alerts the immune system to dying cells. Nature 425:516–21 vity. J Immunol 168:3303–8 Shornick LP, De Togni P, Mariathasan S, Goellner J, Strauss-Schoenberger J, Yamanaka K, Tanaka M, Tsutsui H, Kupper TS, Asahi K, Okamura H et al. Karr RW et al. (1996) Mice deficient in IL-1beta manifest impaired (2000) Skin-specific caspase-1-transgenic mice show cutaneous apop- contact hypersensitivity to trinitrochlorobenzone. J Exp Med 183: tosis and pre-endotoxin shock condition with a high serum level of IL-18. 1427–36 J Immunol 165:997–1003

www.jidonline.org 1963